Substituir a função de janela OVER() pela referência da cláusula WINDOW produz resultados diferentes

Ao preparar uma resposta para outra pergunta aqui , codifiquei uma consulta que continha múltiplas funções de janela com a mesma OVER(...)cláusula. Os resultados foram os esperados.

select ...
    sum(sum(s.price)) over (partition by p.productid
                            order by c.date
                            rows between 6 preceding and current row)
        / nullif(
             sum(count(s.price)) over(partition by p.productid
                                      order by c.date
                                      rows between 6 preceding and current row),
             0)
        as avg7DayPrice
...

Quando tentei substituir as OVER(...)cláusulas duplicadas por uma WINDOWreferência de cláusula compartilhada OVER Last7Days, definida posteriormente como WINDOW Last7Days AS (...), os resultados foram inesperadamente diferentes. (Consulte WINDOWcláusula - nova em 2022.)

select ...
    sum(sum(s.price)) over last7days
        / nullif(sum(count(s.price)) over last7days, 0)
        as avg7DayPrice
...
window last7days as (partition by p.productid
                     order by c.date
                     rows between 6 preceding and current row)

Consulta original completa:

-- Inline window over() clauses
-- Results are as expected
with calendar as (
    select min(date) as date, max(date) as endDate
    from sales
    union all
    select dateadd(day, 1, date), endDate
    from calendar
    where date < enddate
),
products as (
    select distinct productid
    from sales
)
select
    p.productid,
    c.date as salesDate, 
    avg(price) as avg1DayPrice,
    sum(sum(s.price)) over (partition by p.productid
                            order by c.date
                            rows between 6 preceding and current row)
        / nullif(
             sum(count(s.price)) over(partition by p.productid
                                      order by c.date
                                      rows between 6 preceding and current row),
             0)
        as avg7DayPrice
from calendar c
cross join products p
left join sales s
    on s.date = c.date
    and s.productid = p.productid
group by p.productid, c.date

Consulta modificada completa:

-- Reference to common defined window clause
-- Expect same results, but that is not what I get.
with calendar as (
    select min(date) as date, max(date) as endDate
    from sales
    union all
    select dateadd(day, 1, date), endDate
    from calendar
    where date < enddate
),
products as (
    select distinct productid
    from sales
)
select
    p.productid,
    c.date as salesDate, 
    avg(price) as avg1DayPrice,
    sum(sum(s.price)) over last7days
        / nullif(
             sum(count(s.price)) over last7days,
             0)
        as avg7DayPrice
from calendar c
cross join products p
left join sales s
    on s.date = c.date
    and s.productid = p.productid
group by p.productid, c.date
window last7days as (partition by p.productid
                     order by c.date
                     rows between 6 preceding and current row)

Tabela de vendas:

Resultados originais:

Resultados modificados:

Veja este db<>fiddle .

Não entendo por que os resultados são diferentes. Parece que a sum(count(s.price)) over(...)parte do cálculo é 1, 3, 4para a primeira consulta, mas 4, 4, 4na segunda. O plano de execução também é significativamente diferente. Alguém pode explicar isso?